Synthesis, opiate receptor affinity, and conformational parameters of [4

Schiller, Yam, Prosmanne. (7) L. A. Walter and W. K. Chang, J. Med. Chem., 18, 206. (1975). (8) R. C. Cavestri and M. Mokotoff, J. Med. Chem., 20, 149...
1 downloads 0 Views 1MB Size
1110 Journal of Medicinal Chemistry, 1978, Vol. 21, N o . 11 L. A. Walter and W. K. Chang, J . Med. Chem., 18, 206 (1975). R. C. Cavestri and M. Mokotoff, J . Med. Chem., 20, 1493 (1977). S. Shiotani and T. Kometani, Chem. Pharm. Bull., 21, 1053 (1973). S. Shiotani, T. Kometani, K. Mitsuhashi, T. Nozawa, A. Kurobe, and 0. Futsukaichi, J . Med. Chem., 19,803 (1976). S. Shiotani, T. Kometani, and K. Mitsuhashi, J . Med. Chem., 18, 1266 (1975). S. Shiotani, T. Kometani, and K. Mitsuhashi, J . Med. Chem., 20, 310 (1977). K. Mitsuhashi, S. Shiotani, R. Oh-uchi, and K. Shiraki, Chem. Pharm. Bull., 17,434 (1969). M. Takeda and H. Kugita, J . Med. Chem., 13,630 (1970). J. W. Lewis, P. A. Mayor, and D. I. Haddlesey, J . Med. Chem.. 16, 12 (1973). K. E. Opheim and B. M. Cox, J . Med. Chem., 19,867 (1976). P. W. Shiller, C. F. Yam, and M. Lis, Riochemistr3,16,1831 (1977). B. Belleau, T. Conway, F. R. Ahmed, and A. D. Hardy, J . Med. Chem., 17, 907 (1974). T. G. Cochran, J . Med. Chem., 17, 987 (1974). A. P. Feinberg, I. Creese, and S. H. Snyder, Roc. Natl. Acad. Sei. V.S.A., 73, 4215 (1976).

Schiller, Yam, Prosmanne

M. Takeda, H. Inoue, K. Noguchi, Y. Honma, M. Kawamori, G. Tsukamoto, Y. Yamawaki, S. Saito, K. Aoe, T. Date, S. Nurimoto, and G. Hayashi, J . Med. Chem., 20, 221 (1977). P. S. Portoghese, Ace. Chem. Res., 11, 21 (1978). S. Shiotani, T. Kometani, Y. Iitaka, and A. Itai, J . Med. Chem., 21, 153 (1978). G. N.Walker and D. Alkalay, J . Org. Chem., 31,1905 (1966). R. D. Haworth, B. Jones, and T.M.Way, J . Chem. Soc., 10 (1943). S. Shiotani and K. Mitsuhashi, Chem. Pharm. Bull., 14, 324 (1966). M. S. Newman, W. C. Sagar, and C. C. Cochrane, J . Org. Chem., 23, 1832 (1958). K. Nakazawa and S. Baba, Yakugaku Zasshi, 75,378 (1955). A. Kasahara, H. Kojima, Y. Osada, W. Tsukada, and Y. Oshima, Yakugahu Zasshi,88, 606 (1968). J. T. Lichfield, Jr.. and F. Wilcoxon, J . Pharmacol., 96, 99 (1949). F. R. Ahmed and A. D. Hardy, Acta Crystallogr., Sect. H , 31, 2919 (1975). T. G. Cochran and J. E. Abola. Acta Crystallogr.,Sect. H , 31, 919 (1975). Y. Iitaka and A. Itai, unpublished results. A. H. Beckett and A. F. Casy, J . Pharm. Pharmacol., 6,986 (1954).

Synthesis, Opiate Receptor Affinity, and Conformational Parameters of [ 4-Tryptophanlenkephalin Analogues' Peter W. Schiller,* Chun F. Yam, and Janie Prosmanne Laborator3 o f Chemical Biology and Peptide Research, Clinical Research Institute o f Montreal, Montreal, Quebec H2WlR7, Canada Receiued April 26, 1978

A series of analogues of the opioid peptide enkephalin with tryptophan substituted for phenylalanine in position 4 was synthesized by the solid-phase method. The [Trp4]enkephalinanalogues and the corresponding [Phe4]enkephalin analogues displayed nearly parallel affinities in the opiate receptor binding assay throughout the series. In a conformational study fluorescence parameters were measured and intramolecular Tyr-Trp distances were estimated on the basis of resonance energy transfer experiments. No gross conformational differences were observed between analogues with widely differing opiate receptor affinity; however, small but significant changes in the intramolecular distance between the phenol ring and the indole moiety and/or in their relative orientation became apparent in some compounds. Identical intramolecular distances of 9.3 f 0.2 A between the two aromatic rings were obtained with [Trp4,Met5]enkephalin,[Trp4,Leu5]enkephalin,and the N-terminal tetrapeptide comprised in the latter two analogues, indicating the existence of folded conformations in 2 X lo-' M aqueous solution and demonstrating conformational analogy between these three peptides. The conformational parameters are discussed in relation to the observed affinities and the putative opiate receptor topography.

Since the recently discovered opioid peptides [Met5]enkephalin (HSTyr-Gly-Gly-Phe-Met.OH) and [Leu5]enkephalin (HVTyr-Gly-Gly-Phe-LeuSOH)compete with morphine and its derivatives for opiate binding sites,2the enkephalin-opiate receptor system is uniquely suited for studying conformational aspects of a polypeptide-receptor interaction. The multitude of structure-activity relationships established with opiates has provided clues regarding the chemical functions which are critical for binding to the receptor and for activity. On the basis of these studies and taking into account the relatively rigid structure of morphine-derived ligands, descriptions of the opiate receptor topography in terms of binding sites have been put It is thus of considerable interest to demonstrate correspondence between critical chemical functions in morphine derivatives and in enkephalin and to establish their relative spatial disposition in the peptide. In Figure 1 the structural formulas of morphine and a potent derivative of the oripavine family, 7 a - [ l - ( R ) hydroxy-l-methyl-3-phenylpropyl]-6,14-e~~o-ethenotetrahydrooripavine (PEO), are compared with that of [Met6]enkephalin. O r n i ~ s i o nand ~ ~acetylationgJO ~ of the 0022-2623/78/ 1821-1110$01.00/0

a-amino group in enkephalin produce an almost complete loss of activity both in the binding assay and in the guinea pig ileum bioassay. Methylation of the a-amino group to the secondary amine corresponding to the situation in normorphine results in an analogue with still good activity and affinity.'lJ* Finally, N-allylation of the tyrosyl residue induces partial antagonist proper tie^'^ in analogy to the strong antagonism observed with N-allyl derivatives of morphine (e.g., naloxone). Substitution of phenylalanine for tyrosine leads to a drastic reduction in activitys and affinity,14 which demonstrates the importance of the phenolic hydroxyl group for the interaction with the receptor. This finding is in qualitative agreement with the reduced analgesic activity of nonphenolic benzomorphans compared to the corresponding phenolic compound^.'^ 0-Methylation of the tyrosine hydroxyl group engenders a drastic drop in activity" as is the case with 0-methylated derivatives of morphine (e.g., codeine). Clearly, the ensemble of these results supports the idea of a correspondence between the tyrosine moiety in enkephalin and the phenol ring and tertiary nitrogen in morphine-related compounds. Furthermore, these findings suggest similar 1978 American Chemical Society

Journal o f Medicinal Chemistry, 1978, Vol. 21, No. 11 1111

[4- Tryptophanlenkephalin Analogues MORPHINE

ORiPAYlNE "PEO"

0 1

H f r -C H -C-

NH-

II

CH2 - C

II

- N H -C H 2 - C - NHIt

I

CH-

C-

II

N

I

NH-

CH -C - 0

0

II

Figure 1. Structural formulas of morphine, oripavine (PEO), and [Met5]enkephalin. modes of binding4 for morphine and enkephalin. In addition, evidence from a theoretical conformational analysis16 and from an X-ray diffraction study" suggests that the carboxyl terminus of enkephalin and the C-6 hydroxyl (methoxy) group of morphine and its derivatives might have identical functions. Two hypotheses regarding the role of the phenylalanine residue in position 4 have been promulgated. The first proposal is based on studies with analogues in the oripavine family which had shown that attachment of a phenylethyl group to carbon-19 (Figure 1)increases potency by three orders of magnitude relative to m ~ r p h i n e .This ~ observation had led to the proposition of a second lipophilic binding site a t a distance of approximately 10 A from the binding site of the phenol ring. It has been suggested that the phenylalanine residue might correspond to the carbon-19 substituent in oripavine, whereby its correct spatial disposition relative to the tyramine moiety would be brought about by 4 1 or 5 2 hydrogen-bonded ,&bend conformation^'^^^^ or other types of low-energy conformers.16 In an alternative proposal carbon atoms 5 and 6 of the C ring in morphine are considered as an additional binding element which would be provided by the meta and para positions of the phenyl ring of enkephalin in a computer-generated pharmacophoric conformation.20 The latter conformation is characterized by a close proximity of the two aromatic rings (- 5 A) in contrast to the phenol-phenyl distance of about 10 A in the p bend models. The experimental determination of this intramolecular distance is thus of considerable interest. Theoretical conformational analyses produced several folded conformers of low energy, including the 4 1and the 5 2 hydrogen-bonded PI bends,21 the 11'6 bend centered on G l ~ ~ - P hand e ~other , ~ ~ compact structures.16 In the crystalline state the presence of a 6 bend centered on Gly2-Gly3with antiparallel hydrogen bonding between tyrosine and phenylalanine was revealed by X-ray diffractional7 NMR studies indicate the existence of a preferred, rigid backbone conformation both in Me2S0and in aqueous s o l ~ t i o n The . ~ ~results ~ ~ ~ of~ two ~ ~ NMR investigations have been interpreted in favor of orientational freedom of all three side chains relative to the peptide b a c k b ~ n e while , ~ ~ ~a~third ~ study suggests restricted motion for the tyrosyl side chain.24 Some of the conflicting results obtained from NMR experiments have been attributed to pH-dependent conformational changes25and to dimerization a t high concentrations.26 Conformational studies with [Met5]enkephalin and the active analogue [Trp4,Met5]enkephalin have been performed using fluorescence technique^.^^^^^ Fluorescence spectroscopy permits the measurement of conformational parameters in aqueous solution a t low concentrations

-

-

-

( M) where enkephalin presumably exists in the monomeric form. The results of these experiments permitted the following conclusions. (a) Conformers with the tyrosyl hydroxyl engaged in a hydrogen bond are not predominant in aqueous solution. This finding is in contrast to reports22*26 indicating the existence of a hydrogen bond between the phenolic hydroxyl and a backbone carbonyl group in either position 3 or 4. (b) On the basis of a resonance energy transfer experiment, an average intramolecular distance of about 10 A between the two aromatic rings in [Trp4,Met5]enkephalin was determined. (c) Very similar, if not identical conformations or distributions of conformations are observed at pH 1.5 and 5.5. This finding indicates that titration of the C-terminal carboxyl group does not induce a conformational change and provides evidence against the p r o p o ~ a l of ~ ~a, ~ ~ NH,+-COO- interaction as an important factor in stabilizing the conformation a t physiological pH. Fluorescence techniques are of particular value for conformational comparisons in a series of related peptides. In the present paper conformational parameters of enkephalin analogues with widely differing opiate receptor affinity are presented in an attempt to find a possible correlation between solution conformation and receptor binding. Energy Transfer Experiments. In singlet-singlet resonance energy transfer between a donor fluorophore and an acceptor chromophore, the intramolecular donor-acceptor separation, r , is related to the transfer efficiency, E , by eq 1,where the Forster critical distance, Ro, is defined

-

E = R 2 / ( r 6 + Ro6)

(1)

as that donor-acceptor separationzs where 50% of the donor excitation energy is transferred to the acceptor. Ro, measured in centimeters, is related to various measurable spectroscopic parameters by eq 2 , which permits its computation. Ro = [(8.79 x 10-25)(K2/n4)@DoJAD]1'6 (2) K~

= dipole-dipole orientation factor, n = refractive index,

donor fluorescence quantum yield in absence of the acceptor, and Jm = So"FD(X)tA(h)X4 dX = spectral overlap integral between the molar decadic absorption coefficient of the acceptor (eA) and the spectral distribution of the donor fluorescence (FD),normalized to unity and modified by the wavelength factor X4. The spectroscopic properties of phenol (donor) and indole (acceptor) in aqueous solution permit the measurement of intramolecular Tyr-Trp distances in peptides in a range from 8 to 15 A. The transfer efficiency can be determined experimentally by measurement of dDoand the donor fluorescence quantum yield in the presence of the acceptor, dD, on the basis of eq 3. @Do =

E = 1 - (+D/@D') (3) In the case of a peptide with the donor-acceptor pair tyrosine-tryptophan, $Do is best determined with a related peptide containing phenylalanine in place of tryptophan. Phenylalanine, while similar to tryptophan, does not participate in energy transfer at an excitation wavelength of 270 nm. Thus, for the computation of the Tyr-Trp intramolecular distance in [Trp4,Met5]enkephalin(Tyris best determined by measurement Gly-Gly-Trp-Met), bo of the tyrosine fluorescence quantum yield in [Met5]enkephalin [Tyr-Gly-Gly-Phe-Met). For the determination the , tyrosine emission of [Trp4,Met5]enkephalinis of $ J ~ quantified through normalization a t 370 nm of its fluorescence emission spectra obtained with excitation a t

1112 Journal of Medicinal Chemistry, 2978, Vol. 22, N o . 1 1

270 and 293 nm and subsequent subtraction of the normalized spectra. The transfer efficiency can also be determined from the relative enhancement of the acceptor f l u ~ r e s c e n c by e ~ comparison ~~~~ of the fluorescence excitation and absorption spectrum of [Trp4,Met5]enkephalin with the excitation spectrum of [Phe',Trp4,Met5]enkephnlin (Phe-Gly-Gly-Trp-Met). For the transfer of excitation energy from the phenol ring of tyrosine to the indole moiety of tryptophan in [Trp4,Met5]enkephalin,the determination of the transfer efficiency from donor quenching and from the relative enhancement of the acceptor fluorescence yielded values of E = 0.70 f 0.015 and 0.54 f 0.10, respectively." This result indicates that most of the tyrosine fluorescence quenching due to the presence of the acceptor tryptophan is indeed caused by energy transfer. Aside from the distance, the transfer efficiency is also dependent on the relative orientation of the transition dipoles of the donor and acceptor chromophores, which is taken into account by the orientation factor ( x 2 ) in Forster's equation. The implications of the orientation factor on the measurement of intramolecular distances by energy transfer have recently been discussed to great extents3' In the case of dynamic random orientation K' assumes a value of 2 / 3 . The results of a recent t,heoretical study by Haas et al.32indicate that even for fixed orientations the problem of the orientation factor is markedly alleviated if the donor and/or the acceptor fluorophore show mixed polarizations in the wavelength region of spectral overlap. Fluorophores displaying mixed polarizations contain two or three transition dipole moments and their polarizations, p , are low. Thus, the electronic transition of tryptophan in the spectral region of overlap with the tyrosine fluorescence spectrum is characterized by two dipole moments and, therefore, a low polarization ( p = 0.10 f 0.05) is observed. According to the analysis of Haas et al.,32 this value of polarization results in an error of less than 20% for the estimated distance in a situation of fixed orientations if a value of '1 - '> K- - is used. In the case of [Metjlenkephalin, segmental motion relative to the peptide backbone and the existence of three populations of rapidly interconverting rotamers has been demonstrated for the aromatic side chains in positions 1 and 4 by NMR s p e c t r o ~ c o p y . ~Obviously, ~,~~ this orientational freedom of the fluorophores reduces the error in the distance even further. While energy transfer experiments provide a valuable tool to estimate absolute intramolecular distances with an accuracy of about 2 0 % , their potential in studies aiming at conformational comparisons is perhaps of greater interest. Transfer efficiencies in conjunction with fluorescence quantum yields represent sensitive parameters permitting the detection of conformational differences in a series of peptide analogues containing the same donor-acceptor pair. Among the various parameters contributing to Ro, $Do can assume different values in a series are usually constant in a of analogues, while n, and JAD given solvent. In the case where the same values of +no are observed for two analogues, the transfer efficiency in both compounds is only a function of r and K~ (eq 4,where E = So6K2/(r6+ S o 6 ~ 2 ) (4) So = [(8.79 X 1 0 ~ 2 ' ) ( l / ~ 4 ) ~ ~ 0 J=A constant. ~ ] 1 / f i If, in addition, identical values of E are also observed in the same two analogues, this is a strong indication for an analogous spatial disposition of the donor and acceptor fluorophores in the two peptides. The possibility that different values of r and K~ in each peptide compensate each other to produce exactly the same value of E is extremely unlikely.

Schiller, Yam, Prosmanne Table I. Opiate Receptor Affinities of Enkephalin Analoguesu __ re1 receptor affinities, O/n ([Mets]enkephalin = 100%)

I I1 I11 IV V VI VI1 a

analogue Tyr-Gly-Gly-X-Met Tyr-Gly-Gly-X-Leu Tyr-Gly-Gly-X Tyr-Gly-Gly-X-Met-Thr Tyr-D -Ala-Gly-X-Met Tyr-L -Ala-Gly-X-Met Tyr-Gly-Gly-D-X-Met Mean of three determinations

X = Phe 100 38 F 6 2 x 0 2 84 + 16 54 ?- 1 4 5 t 2